基于改进灰狼算法和自适应分裂KD-Tree的点云配准方法  

Point Cloud Registration Method Based on Improved Grey Wolf Algorithm and Adaptive Splitting KD-Tree

作　　者：杜沅昊 耿秀丽[1,2] 徐诚智 刘银华[3] Du Yuanhao;Geng Xiuli;Xu Chengzhi;Liu Yinhua(Business School,University of Shanghai for Science and Technology,Shanghai 200093,China;School of Intelligent Emergency Management,University of Shanghai for Science and Technology,Shanghai 200093,China;School of Mechanical Engineering,University ofShanghai for Science and Technology,Shanghai 200093,China)

机构地区：[1]上海理工大学管理学院,上海200093 [2]上海理工大学智慧应急管理学院,上海200093 [3]上海理工大学机械工程学院,上海200093

出　　处：《系统仿真学报》2025年第2期424-435,共12页Journal of System Simulation

基　　金：国家自然科学基金(72271164);教育部人文社会科学研究规划基金(19YJA630021);上海市浦江人才计划(22PJD048)。

摘　　要：针对传统GWO存在搜索效率不足、易陷入局部最优等问题,提出了一种基于改进GWO和迭代最近点(ICP)的工业复杂零件点云配准方法。针对GWO随机初始化导致种群分布不均匀的问题,采用混沌映射对灰狼种群进行初始化,使种群更加均匀地分布在搜索空间内;引入一种非线性控制参数策略,平衡灰狼算法的局部搜索和全局搜索能力;融合精英反向学习,提高算法后期解的质量;利用ICP算法进行精配准。设计一种自适应分裂维度的方法,动态选择分裂维度,提高点云数据质量。仿真结果表明:IGWO相较于3种对比算法的RMSE平均提高了80.31%、73.99%、47.7%。Traditional GWO algorithms suffer from limitations such as insufficient search efficiency and susceptibility to local optima.A novel method for the registration of point clouds of complex industrial components is proposed based on an improved GWO algorithm and ICP.To address the problem of uneven population distribution caused by random initialization in GWO,chaotic mapping is employed to initialize the gray wolf population,ensuring a more uniform distribution of individuals within the search space.A non-linear control parameter strategy is introduced to strike a balance between the algorithm's local search and global search capabilities.Elite reverse learning is integrated to improve the quality of the algorithm's solutions.The refined registration is achieved using the ICP algorithm.An adaptive dimension splitting method is developed.This method dynamically selects the splitting dimensions to enhance the quality of the point cloud data.The experiments show that the RMSE of IGWO increases by 80.31%,73.99%and 47.7%on average compared with the other three comparison algorithms.

关 键 词：改进灰狼算法 混沌映射 非线性参数 精英反向学习 点云配准 自适应分裂维度 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP391.41[自动化与计算机技术—控制科学与工程]